1 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
2 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +bf16 -msve-vector-bits=512 -fallow-half-arguments-and-returns -S -disable-llvm-passes -emit-llvm -o - %s | FileCheck %s
3 
4 #include <arm_sve.h>
5 
6 #define N __ARM_FEATURE_SVE_BITS
7 
8 typedef svint32_t fixed_int32_t __attribute__((arm_sve_vector_bits(N)));
9 typedef svbool_t fixed_bool_t __attribute__((arm_sve_vector_bits(N)));
10 
11 fixed_bool_t global_pred;
12 fixed_int32_t global_vec;
13 
14 // CHECK-LABEL: @foo(
15 // CHECK-NEXT:  entry:
16 // CHECK-NEXT:    [[RETVAL:%.*]] = alloca <16 x i32>, align 16
17 // CHECK-NEXT:    [[PRED_ADDR:%.*]] = alloca <vscale x 16 x i1>, align 2
18 // CHECK-NEXT:    [[VEC_ADDR:%.*]] = alloca <vscale x 4 x i32>, align 16
19 // CHECK-NEXT:    [[PG:%.*]] = alloca <vscale x 16 x i1>, align 2
20 // CHECK-NEXT:    store <vscale x 16 x i1> [[PRED:%.*]], <vscale x 16 x i1>* [[PRED_ADDR]], align 2
21 // CHECK-NEXT:    store <vscale x 4 x i32> [[VEC:%.*]], <vscale x 4 x i32>* [[VEC_ADDR]], align 16
22 // CHECK-NEXT:    [[TMP0:%.*]] = load <vscale x 16 x i1>, <vscale x 16 x i1>* [[PRED_ADDR]], align 2
23 // CHECK-NEXT:    [[TMP1:%.*]] = load <8 x i8>, <8 x i8>* @global_pred, align 2
24 // CHECK-NEXT:    [[TMP2:%.*]] = load <vscale x 16 x i1>, <vscale x 16 x i1>* bitcast (<8 x i8>* @global_pred to <vscale x 16 x i1>*), align 2
25 // CHECK-NEXT:    [[TMP3:%.*]] = load <8 x i8>, <8 x i8>* @global_pred, align 2
26 // CHECK-NEXT:    [[TMP4:%.*]] = load <vscale x 16 x i1>, <vscale x 16 x i1>* bitcast (<8 x i8>* @global_pred to <vscale x 16 x i1>*), align 2
27 // CHECK-NEXT:    [[TMP5:%.*]] = call <vscale x 16 x i1> @llvm.aarch64.sve.and.z.nxv16i1(<vscale x 16 x i1> [[TMP0]], <vscale x 16 x i1> [[TMP2]], <vscale x 16 x i1> [[TMP4]])
28 // CHECK-NEXT:    store <vscale x 16 x i1> [[TMP5]], <vscale x 16 x i1>* [[PG]], align 2
29 // CHECK-NEXT:    [[TMP6:%.*]] = load <vscale x 16 x i1>, <vscale x 16 x i1>* [[PG]], align 2
30 // CHECK-NEXT:    [[TMP7:%.*]] = load <16 x i32>, <16 x i32>* @global_vec, align 16
31 // CHECK-NEXT:    [[CASTSCALABLESVE:%.*]] = call <vscale x 4 x i32> @llvm.experimental.vector.insert.nxv4i32.v16i32(<vscale x 4 x i32> undef, <16 x i32> [[TMP7]], i64 0)
32 // CHECK-NEXT:    [[TMP8:%.*]] = load <vscale x 4 x i32>, <vscale x 4 x i32>* [[VEC_ADDR]], align 16
33 // CHECK-NEXT:    [[TMP9:%.*]] = call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[TMP6]])
34 // CHECK-NEXT:    [[TMP10:%.*]] = call <vscale x 4 x i32> @llvm.aarch64.sve.add.nxv4i32(<vscale x 4 x i1> [[TMP9]], <vscale x 4 x i32> [[CASTSCALABLESVE]], <vscale x 4 x i32> [[TMP8]])
35 // CHECK-NEXT:    [[CASTFIXEDSVE:%.*]] = call <16 x i32> @llvm.experimental.vector.extract.v16i32.nxv4i32(<vscale x 4 x i32> [[TMP10]], i64 0)
36 // CHECK-NEXT:    store <16 x i32> [[CASTFIXEDSVE]], <16 x i32>* [[RETVAL]], align 16
37 // CHECK-NEXT:    [[TMP11:%.*]] = load <16 x i32>, <16 x i32>* [[RETVAL]], align 16
38 // CHECK-NEXT:    [[CASTSCALABLESVE1:%.*]] = call <vscale x 4 x i32> @llvm.experimental.vector.insert.nxv4i32.v16i32(<vscale x 4 x i32> undef, <16 x i32> [[TMP11]], i64 0)
39 // CHECK-NEXT:    ret <vscale x 4 x i32> [[CASTSCALABLESVE1]]
40 //
41 fixed_int32_t foo(svbool_t pred, svint32_t vec) {
42   svbool_t pg = svand_z(pred, global_pred, global_pred);
43   return svadd_m(pg, global_vec, vec);
44 }
45 
46 // CHECK-LABEL: @test_ptr_to_global(
47 // CHECK-NEXT:  entry:
48 // CHECK-NEXT:    [[RETVAL:%.*]] = alloca <16 x i32>, align 16
49 // CHECK-NEXT:    [[GLOBAL_VEC_PTR:%.*]] = alloca <16 x i32>*, align 8
50 // CHECK-NEXT:    store <16 x i32>* @global_vec, <16 x i32>** [[GLOBAL_VEC_PTR]], align 8
51 // CHECK-NEXT:    [[TMP0:%.*]] = load <16 x i32>*, <16 x i32>** [[GLOBAL_VEC_PTR]], align 8
52 // CHECK-NEXT:    [[TMP1:%.*]] = load <16 x i32>, <16 x i32>* [[TMP0]], align 16
53 // CHECK-NEXT:    store <16 x i32> [[TMP1]], <16 x i32>* [[RETVAL]], align 16
54 // CHECK-NEXT:    [[TMP2:%.*]] = load <16 x i32>, <16 x i32>* [[RETVAL]], align 16
55 // CHECK-NEXT:    [[CASTSCALABLESVE:%.*]] = call <vscale x 4 x i32> @llvm.experimental.vector.insert.nxv4i32.v16i32(<vscale x 4 x i32> undef, <16 x i32> [[TMP2]], i64 0)
56 // CHECK-NEXT:    ret <vscale x 4 x i32> [[CASTSCALABLESVE]]
57 //
58 fixed_int32_t test_ptr_to_global() {
59   fixed_int32_t *global_vec_ptr;
60   global_vec_ptr = &global_vec;
61   return *global_vec_ptr;
62 }
63 
64 //
65 // Test casting pointer from fixed-length array to scalable vector.
66 // CHECK-LABEL: @array_arg(
67 // CHECK-NEXT:  entry:
68 // CHECK-NEXT:    [[RETVAL:%.*]] = alloca <16 x i32>, align 16
69 // CHECK-NEXT:    [[ARR_ADDR:%.*]] = alloca <16 x i32>*, align 8
70 // CHECK-NEXT:    store <16 x i32>* [[ARR:%.*]], <16 x i32>** [[ARR_ADDR]], align 8
71 // CHECK-NEXT:    [[TMP0:%.*]] = load <16 x i32>*, <16 x i32>** [[ARR_ADDR]], align 8
72 // CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds <16 x i32>, <16 x i32>* [[TMP0]], i64 0
73 // CHECK-NEXT:    [[TMP1:%.*]] = load <16 x i32>, <16 x i32>* [[ARRAYIDX]], align 16
74 // CHECK-NEXT:    store <16 x i32> [[TMP1]], <16 x i32>* [[RETVAL]], align 16
75 // CHECK-NEXT:    [[TMP2:%.*]] = load <16 x i32>, <16 x i32>* [[RETVAL]], align 16
76 // CHECK-NEXT:    [[CASTSCALABLESVE:%.*]] = call <vscale x 4 x i32> @llvm.experimental.vector.insert.nxv4i32.v16i32(<vscale x 4 x i32> undef, <16 x i32> [[TMP2]], i64 0)
77 // CHECK-NEXT:    ret <vscale x 4 x i32> [[CASTSCALABLESVE]]
78 //
79 fixed_int32_t array_arg(fixed_int32_t arr[]) {
80   return arr[0];
81 }
82 
83 // CHECK-LABEL: @address_of_array_idx(
84 // CHECK-NEXT:  entry:
85 // CHECK-NEXT:    [[RETVAL:%.*]] = alloca <8 x i8>, align 2
86 // CHECK-NEXT:    [[ARR:%.*]] = alloca [3 x <8 x i8>], align 2
87 // CHECK-NEXT:    [[PARR:%.*]] = alloca <8 x i8>*, align 8
88 // CHECK-NEXT:    [[RETVAL_COERCE:%.*]] = alloca <vscale x 16 x i1>, align 16
89 // CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [3 x <8 x i8>], [3 x <8 x i8>]* [[ARR]], i64 0, i64 0
90 // CHECK-NEXT:    store <8 x i8>* [[ARRAYIDX]], <8 x i8>** [[PARR]], align 8
91 // CHECK-NEXT:    [[TMP0:%.*]] = load <8 x i8>*, <8 x i8>** [[PARR]], align 8
92 // CHECK-NEXT:    [[TMP1:%.*]] = load <8 x i8>, <8 x i8>* [[TMP0]], align 2
93 // CHECK-NEXT:    store <8 x i8> [[TMP1]], <8 x i8>* [[RETVAL]], align 2
94 // CHECK-NEXT:    [[TMP2:%.*]] = bitcast <vscale x 16 x i1>* [[RETVAL_COERCE]] to i8*
95 // CHECK-NEXT:    [[TMP3:%.*]] = bitcast <8 x i8>* [[RETVAL]] to i8*
96 // CHECK-NEXT:    call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[TMP2]], i8* align 2 [[TMP3]], i64 8, i1 false)
97 // CHECK-NEXT:    [[TMP4:%.*]] = load <vscale x 16 x i1>, <vscale x 16 x i1>* [[RETVAL_COERCE]], align 16
98 // CHECK-NEXT:    ret <vscale x 16 x i1> [[TMP4]]
99 //
100 fixed_bool_t address_of_array_idx() {
101   fixed_bool_t arr[3];
102   fixed_bool_t *parr;
103   parr = &arr[0];
104   return *parr;
105 }
106